We request support for the continued statistical analyses of the data recently acquired by a major long-term follow-up of chronic mental patients in the community. 269 Vermont State Hospital patients, most with diagnoses of schizophrenia, some with primary affective disorders, and other psychiatric illnesses were originally selected for chronic disability, rehabilitated, and released to the community in the mid-1950's. 97% of this group has been located and/or accounted for 20 years post index discharge but our catamnesis years range from 21 to 58 years. Two field interviews have been conducted with each live proband (N=191). The first interview included a multivariate cross-sectional assessment of outcome. The second interview included a Meyer/Leighton Life Chart and contributed a longitudinal documentation of patterns, shifts, and trends in the course for members of the cohort. These patterns were derived from a multidimensional year-by-year follow-along and follow-back over the 20-year period. Informants who knew the subjects well were interviewed to verify current status and historical data. The Vermont Community Questionnaire instrument battery (VCQ) contained an innovative weaving of several new and classic scales and schedules which obtained structured measures of outcome. Hospital and Vocational Rehabilitation records were reviewed with the Hospital Record Review Battery (HRRF) compiled from a modification of W.H.O.'s Psychiatric and Personal History Form and Strauss' Case Record Review. The first of the major statistical cross-sectional analyses was completed on the live DSM I schizophrenic probands (n=149), 82 subjects who have been rediagnosed as having met the DSM III schizophrenic criteria at selection in the mid-50's and the deceased subjects. These analyses on both live and deceased samples revealed over half recovered or significantly improved. Outcome was found to be heterogeneous across domains of functioning. Many important questions are now being addressed, utilizing the year-by-year longitudinal follow-back information as significant and modifying variables to assess current status within and between domains of functioning. Specific aims are: 1) to identify predictors which might differentiate long-term outcome for patients selected for their chronicity; 2) to delineate the long-term patterns, shifts, and trends; and 3) to produce a DSM III casebook on the long-term course of schizophrenia highlighting post morbid and recompensation processes. Complex multivariate statistical strategies have been described.